LAUSR

Abstract Self-citation is a common practice in soil science publications, but can we find a common behaviour in soil science papers, and are there outliers? Here we investigate self-referencing (referencing earlier papers on which a researcher is an author or co-author) for papers published in nine leading soil science journals (Soil Biology and Biochemistry, Geoderma, Soil & Tillage Research, Catena, European Journal of Soil Science, Soil Science Society of America Journal, Soil Use and Management, Geoderma Regional, and SOIL) in 2018 to understand the self-referencing distribution of soil science papers. From 1753 research and review papers, a typical soil science paper (based on median values) had 5 authors, 56 references, and 10 pages. Interestingly, these numbers were very similar across all nine journals. We found that the median rate of self-referencing was 10%, and an interquartile range from 5 to 17%. The values were also very similar across the nine journals. The self-referencing values were reasonable and comparable to other disciplines and a past study of pedometrics papers. Self-referencing rates increase with the number of references and authors. However, its distribution was quite variable but tended to be less than 25% for papers with less than 10 authors with less than 60 references. We further conducted a graphical procedure on the self-referencing distribution to establish outlier thresholds related to the inter-quartile range. We found thresholds of self-referencing between 35% and 53%, which would suggest high values. There were only 61 papers (3.5% of total) above the lower threshold. We also established median and 80-percentile models based on the number of authors and the number of references. Based on the data, we can present an acceptable range of self-references, followed by an expert interpretation of the citation behaviour. This way can help authors, editors and reviewers to understand whether exceptional citation is at work.